Datasets:

rmems
/

multi-agent-ouroboros-swarm-grok46

source_file string	source_line int64	record_json string
data/raw/batch-r01.jsonl	1	{"id":"maos-fh-r01-001","title":"ZINCFELL HDG-5: air-knife r_dP 4.8 kPa beats coat.mean.in_band by 188 us; correct MODIFY still loses 1.44 km of edge-heavy GI to a pre-t0 dross plug","rights":{"provider":"SpaceXAI/xAI","model":"grok-4.6","channel":"consumer","subscription_plan":"SuperGrok Heavy","generation_surface":"S...
data/raw/batch-r02.jsonl	1	{"id":"maos-r02-001","title":"SLUICE-HEARTH CU-2: notch IR 1180 C beats metal.in_band by 192 us; correct MODIFY still punches two tuyeres after a pre-t0 slag-sluice freeze","rights":{"provider":"SpaceXAI/xAI","model":"grok-4.6","channel":"consumer","subscription_plan":"SuperGrok Heavy","generation_surface":"SuperGrok H...
data/raw/batch-r02c.jsonl	1	{"id": "maos-r02c-001", "title": "FROSTQUAY CS-9: lidar.foul 42 mm beats fleet.in_band by 216 us; correct MODIFY still drops aisle 7 after a pre-t0 rack sag", "rights": {"provider": "SpaceXAI/xAI", "model": "grok-4.6", "channel": "consumer", "subscription_plan": "SuperGrok Heavy", "generation_surface": "SuperGrok Heavy...
data/raw/batch-r03.jsonl	1	{"id":"maos-r03-001","title":"QUAY-FEN QF-AY-3: cell-47 IR 94.2 C beats string-mean-in-band by 192 us; correct MODIFY still loses an O2-header knock-out to a pre-t0 separator tear","rights":{"provider":"SpaceXAI/xAI","model":"grok-4.6","channel":"consumer","subscription_plan":"SuperGrok Heavy","generation_surface":"Sup...
data/raw/batch-r04.jsonl	1	{"id":"maos-r04-001","title":"LOCKFEN PS-4: screen dP 18.4 kPa beats drain-level-in-band by 204 us; correct MODIFY still loses 180 m of peat berm to a pre-t0 intake slip","rights":{"provider":"SpaceXAI/xAI","model":"grok-4.6","channel":"consumer","subscription_plan":"SuperGrok Heavy","generation_surface":"SuperGrok Hea...
data/raw/batch-r21.jsonl	1	"{\"id\":\"maos-r21-001\",\"title\":\"GLIMMERAXLE CAV-7: radar residual 18.4 dBsm beats fuse.gap_ok (...TRUNCATED)
data/raw/batch-r22.jsonl	1	"{\"id\":\"maos-r22-001\",\"title\":\"PRILLGHYLL UT-4: orifice-27 IR 168 C beats header-in-band by 1(...TRUNCATED)
data/raw/batch-r23.jsonl	1	"{\"id\":\"maos-r23-001\",\"title\":\"SPIT-LOCK AT-6: nozzle-4 steam TC 612.4 C beats header-steam-i(...TRUNCATED)
data/raw/batch-r23c.jsonl	1	"{\"id\":\"maos-r23c-001\",\"title\":\"MYCOSTAITH CT-7: Z-6 local TC 41.2 C beats tunnel-mean-in-ban(...TRUNCATED)
data/raw/batch-r41.jsonl	1	"{\"id\":\"maos-r41-001\",\"title\":\"HOLLOWMERE PX-5: harvest-mass residual 6.8 pct beats do.in_ban(...TRUNCATED)

End of preview. Expand in Data Studio

Multi-Agent Ouroboros Swarm (Grok 4.6)

Rights & intended use: public research corpus, not training data. Hosted frontier-model outputs are research-only inputs under project policy (synthetic-factory#161): intended_use: research_only, project_training_policy: blocked. Not training data for any model-weight update. Machine-readable record: rights.json. License: Synthetic Factory Research-Only License v1.0 (license: other, see LICENSE) (non-commercial).

Release status: the raw, uncurated payload of the 2026-09-02 SuperGrok Heavy run is published under data/raw/. It is available for inspection, evaluation, and reproducibility, but it is not training-ready (release-status.json: release_stage: raw_uncurated_public, training_ready: false, strict_audit: pending_curation_integration).

Visibility: public raw-data repository.

Rights and intended use

This is a public research corpus, not training data.

Under the Synthetic Factory project-policy decision recorded in rmems/synthetic-factory#161, outputs of hosted frontier models (here: Grok 4.6 (xAI), generated in the SuperGrok Heavy consumer chat surface) are research-only inputs:

intended_use: research_only
project_training_policy: blocked

These records must not enter SFT, DPO, RL, distillation, continued pretraining, or any other model-weight update in this project. training_ready: false is a data-quality statement; even a future curated release would not make this corpus eligible for weight updates under project policy. That policy value does not change merely because a provider document later permits training — changing it would require a separate recorded project decision.

Research retention, evaluation, and redistribution rights are tracked separately per provider/channel/date. The owner's decision on #161 (2026-09-04) is that research-only outputs may be retained, evaluated, and redistributed for research while training stays blocked; the machine-readable statuses in rights.json nevertheless fail closed as unresolved until the provider terms snapshot is pinned (synthetic-factory#163). provider_training_status and weight_publication_status are unresolved, which resolves to blocked for any dependent activity.

This public raw release is licensed under Synthetic Factory Research-Only License v1.0 (license: other, see LICENSE). That research-only license grants reuse permissions; it does not make the records training-ready, and it does not lift the project's training block.

What this dataset is

Synthetic safety-gate adjudication trajectories staged inside multi-agent scenarios. Kind comes from the payload, not the slug: every published JSONL row is a thalamic-gate wrap record, not a standalone multi-agent conversation transcript or a homogeneous swarm corpus. In every record a proposed action is put to a relay gate that returns ACCEPT, MODIFY, or REJECT, and the record carries the executed action, its downstream outcome, and a reward decomposition. The unit of every published record is the gate decision, not the swarm dialogue. The per-round swarm transcripts (swarm-transcript-rNN.md) that the factory wrote next to each batch are not part of this release.

The release contains 21 thalamic-gate wrap records across 21 JSONL files (~0.74 MB of raw JSONL), all stamped meta.factory: multi-agent-ouroboros-swarm, meta.generator: grok-4.6, and meta.run_label: 2026-09-02-final-heavy. Record ids run maos-fh-r01-001 … maos-r68-001 and are unique across the release. Gate outcomes: MODIFY 21. Every row shares the top-level keys executed_action, future_outcome, gate_snn, id, meta, proposed_action, raster, reward_components, rights, safety_decision, spike_events, state, title. Some rows add provenance.

This is the Grok 4.6 counterpart of rmems/multi-agent-ouroboros-swarm (Claude Fable 5). The two datasets come from the same factory prompts and round protocol but from different generators and different runs; they live in separate repositories and separate collections and are never merged.

Provenance

Generator: xAI Grok 4.6, run in the SuperGrok Heavy consumer chat surface (meta.rights.subscription_plan: SuperGrok Heavy), final run 2026-09-02 (meta.run_label: 2026-09-02-final-heavy). Every record carries meta.generator: grok-4.6 and a per-record meta.rights block recording the provider, model, channel, plan, surface, generation timestamp, and record-level policy fields (21 of 21 rows also carry the individual research/redistribution status fields). Record-level generation_surface values in this release: SuperGrok Heavy chat: 21.
Method: Synthetic Data Factory prompts run in SuperGrok Heavy chat and published through the factory's round protocol; operator and review: Raul Montoya Cardenas; release engineering for this upload: Claude Fable 5.1 (Claude Code). Details in provenance.json.
Rounds published: r01, r02, r02c, r03, r04, r21, r22, r23, r23c, r41, r42, r43, r61, r62, r63, r63c, r64, r65, r66, r67, r68.
Selection rule: batch-rNNc.jsonl files are additional batches of the same round; both are published. The only duplicate, batch-r21.jsonl of failure-as-fuel-preference-cascade (identical ids to batch-r21c.jsonl), is omitted. Every NOTES-r*.md written by the factory is published verbatim under notes/.
Source: the factory's immutable raw tree outputs/raw/2026-09-02-final-heavy/multi-agent-ouroboros-swarm/. Round markers, reservation and completion markers, NEXT_ROUND.json, hidden files, and staging directories are not part of the release.
Gates: pipelines/validate_run.py and pipelines/check_records.py from rmems/synthetic-factory report 0 errors on exactly the published files; rights.json is accepted by pipelines/rights_document.py (schema 0.1.0).

Raw file	Records	SHA-256 (prefix)
`data/raw/batch-r01.jsonl`	1	`a0b9a9c7c6a5…`
`data/raw/batch-r02.jsonl`	1	`90f1ab7902f4…`
`data/raw/batch-r02c.jsonl`	1	`75bf264fbe47…`
`data/raw/batch-r03.jsonl`	1	`a2d89a83fd12…`
`data/raw/batch-r04.jsonl`	1	`213ce4a7bf89…`
`data/raw/batch-r21.jsonl`	1	`5a6b32c1e44e…`
`data/raw/batch-r22.jsonl`	1	`443211d25f93…`
`data/raw/batch-r23.jsonl`	1	`8f826e6aeeda…`
`data/raw/batch-r23c.jsonl`	1	`99d5af2203be…`
`data/raw/batch-r41.jsonl`	1	`b4b70f96692b…`
`data/raw/batch-r42.jsonl`	1	`c7f0640ba035…`
`data/raw/batch-r43.jsonl`	1	`2969e268d01c…`
`data/raw/batch-r61.jsonl`	1	`b4c184bc50bc…`
`data/raw/batch-r62.jsonl`	1	`4a20ce7a042d…`
`data/raw/batch-r63.jsonl`	1	`aca32d47b351…`
`data/raw/batch-r63c.jsonl`	1	`e4a65c3cc110…`
`data/raw/batch-r64.jsonl`	1	`23ea9c353f29…`
`data/raw/batch-r65.jsonl`	1	`a75012e6cc5e…`
`data/raw/batch-r66.jsonl`	1	`179d39fba377…`
`data/raw/batch-r67.jsonl`	1	`811e5f11286f…`
`data/raw/batch-r68.jsonl`	1	`46a78c1ed305…`

Full digests are in provenance.json raw_snapshot.

How to read

data/raw/batch-r*.jsonl is the source of truth: each file is a byte-for-byte copy of the selected factory file, one JSON record per line. Do not rewrite it; do not point a default config at data/raw/*.jsonl (nested key-bags drift across rounds).
data/viewer/records.parquet is a lossless viewer projection with three columns: source_file (the repo-relative raw path), source_line (1-based physical line) and record_json (the verbatim JSON line). Joining a file's record_json values in source_line order with \n reproduces the raw file byte for byte. The Dataset Viewer config viewer / split train indexes 21 rows.
notes/NOTES-r*.md are the factory's per-round notes, verbatim.
rights.json, release-status.json, and provenance.json are the machine-readable rights, release, and provenance sidecars.

import json
import pyarrow.parquet as pq

table = pq.read_table("data/viewer/records.parquet")
record = json.loads(table.column("record_json")[0].as_py())
print(record["id"], record["meta"]["generator"], record["meta"]["round"])

Do not train on any field of this raw Hub copy; it is evidence for research, evaluation, and cross-generator comparison only.

Relationship to the Fable dataset

rmems/multi-agent-ouroboros-swarm holds the Claude Fable 5 rounds of the same factory; this repository holds the Grok 4.6 rounds of the 2026-09-02 SuperGrok Heavy run. Same family, same schema lineage, different generator, different run, different license (the Fable release is Apache-2.0; this one is Synthetic Factory Research-Only License v1.0 (license: other, see LICENSE)). Grok output is never pushed into the Fable repository and the two are never merged; comparing them is one of the intended research uses.

License

This public raw release is licensed under the Synthetic Factory Research-Only License v1.0 (license: other, license_name: synthetic-factory-research-only) (Synthetic Factory Research-Only License v1.0 (license: other, see LICENSE)). That license grants non-commercial reuse permissions; it does not make the records training-ready or factual real-world measurements, and project policy blocks training use regardless of license.